1. Field of the Invention
The present invention relates to a radio frequency circuit which controls the output power of a radio frequency signal in power amplification and a communication system equipped with that radio frequency circuit.
2. Description of the Related Art
In view of reduction in power consumption, transmission power is controlled in many radio communication systems including a mobile communication system. However, when the output power is lowered, degradation of the power addition efficiency η add occurs in the main power amplifier and a driver amplifier located at a preceding stage of the main power amplifier. The power addition efficiency η add can be obtained as a result of dividing a difference between the RF input power Prfin and the RF output power Prfout by the DC input power Pdcin. That is, η add=(Prfout−Prfin)/Pdcin.
FIG. 15A shows a conventional radio frequency circuit which can compensate for degradation of the power addition efficiency. That is, the bias voltage is adjusted to lower the output power, appropriately. For example, when the output power is lowered to ¼, the bias voltage is changed from B2 to B1 as shown in FIG. 15B. However, the radio frequency circuit requires a variable voltage source V1 capable of providing a large current. In general, a DC-DC converter shown in FIG. 15C is used as the variable voltage source V1, thus increasing the number of components, the mounting area thereof, and the cost. Further, in this ratio frequency circuit, it difficult to maintain the efficiency because the proportion of the knee voltage to the bias voltage increases as the output power drops. Moreover, it is difficult to maintain a high efficiency over a wide range of output power because the minimum output voltage of the variable voltage source V1 is of the order of 1 V in association with the reference voltage of the internal regulator and the efficiency of the DC-DC converter drops at low output voltages. In the radio frequency circuit, a variable voltage source V2 is also used for controlling the bias current, which becomes an idle current at zero signal input. However, the bias current is adjusted such that the power gain of the power amplifier remains unchanged as much as possible even at low output levels because the radio frequency circuit is provided for the purpose of keeping the power addition efficiency from decreasing with decreasing output power.
A method, in which a fixed voltage source is used instead of the variable voltage source V1 shown in FIG. 15A, has been also proposed. With the method, only the bias current is controlled by the variable voltage source V2 under the fixed bias voltage. However, the bias current is adjusted such that the power gain of the power amplifier remains unchanged as much as possible even at low output levels because the radio frequency circuit of this method is also provided for the purpose of keeping the power addition efficiency from decreasing with decreasing output power. For example, when the power amplifier includes two stages of amplification to attain the power gain of about 24 dB at maximum output, the bias point is adjusted such that the power gain is maintained at about 21 dB at low output levels as well.
The principle of the conventional method that controls only the bias current based on the variable voltage source V2 will be described. FIG. 8 shows the relationship between the output power of the power amplifier and the ratio of an undesired signal component to a desired signal component (hereinafter referred to as the U/D ratio) in an RF signal output from the power amplifier. The undesired signal is produced due to non-linearity of the power amplifier. Generally, the voltage source V2 is fixed to a voltage value causing that the U/D ratio shifts with a constant margin with respect to the reference value R upon an increase of the output power of the power amplifier and exceeds the reference value R at maximum output power P as shown in FIG. 8. When the amplifier bias point of V2 is set to a value nearer the class B area than the fixed value, the characteristic curve of the U/D ratio to the output power as a whole will move upward in FIG. 8 with a slight change in shape. If a desired output power is lower than the maximum output power P, the power addition efficiency can be compensated for by adjusting the bias point of V2 to a value near the class B area in a range that the U/D ratio to the output power as a whole does not exceed the reference value R.
A radio frequency circuit shown in FIG. 14 is uses as another means for compensating for deterioration of the power addition efficiency. With this radio frequency circuit, power amplifiers different in maximum output power, AMP1 (maximum output power: −20 dBmW), AMP2 (5 dBmW) and AMP3 (3 dBmW), are arranged in series via switch circuits S71, S72, S73 and S74, and bypass circuits 75 and 76 are arranged in parallel with the power amplifiers AMP2 and AMP3, respectively. The switch circuits S71, S72, S73 and S74 perform switching between the power amplifier AMP2 and bypass circuit 75 and between the amplifier AMP3 and bypass circuit 76 so as to change the configuration of signal paths behind the power amplifier AMP1.
This radio frequency circuit operates to disconnect unnecessary power amplifiers at low output power levels and reduce the idle current dissipated by the power amplifiers. Assume that the switches S71 to S74 and the bypass circuits 75 and 76 are free from loss. To provide output power higher than 5 dBmW, the amplifiers AMP1, AMP2, and AMP3 will be required. However, the amplifier AMP3 may be unnecessary and disconnected when the output power is not higher than 5 dBmW. Likewise, both the amplifiers AMP2 and AMP3 can be unnecessary and disconnected when the output power is not higher than −20 dBmW. Thus, the idle current dissipated by at least one of the amplifiers AMP2 and AMP3 can be reduced when the output power is below 5 dBmW.
However, although the switch-based method as shown in FIG. 14 is straightforward, the insertion loss of each of the switches becomes a problem. In particular, the insertion loss of the switch S74 at the output of the final-stage amplifier AMP3 greatly reduces the power addition efficiency. Assuming the insertion loss of each switch to be 1 dB, the output power of the amplifier AMP3 must be 31 dBmW. If the power addition efficiency of the amplifier AMP3 is 40% at maximum output and the supply voltage and the power gain of the amplifier AMP3 are 3.5 V and 25 dB, the consumption current will be 896 mA. In the absence of the switch, the radio frequency circuit has only to produce an output power of 30 dBmW. The consumption current will be 712 mA in the case where the power addition efficiency is 40%, the supply voltage is 3.5 V and the power gain is 25 dB. The insertion of a switch having an insertion loss of 1 dB at the output of the amplifier AMP3 results in an increase of consumption current by 1.26 times. This is nothing else but to increase the device size by 1.26 times in order to increase the output power by 1 dB. The consumption current is increased by 1.26 times over the whole range of output power. Naturally, the idle current is also increased by 1. 26 times. To compensate for 1 dB of insertion loss of each of the switches S72 and S73 between the amplifiers AMP2 and AMP3, the amplifier AMP2 is required to compensate for a total of 3 dB including the insertion loss of the switch S74. In many cases, more stringent distortion criteria are imposed on the amplifier AMP2 than on the amplifier AMP3 and, to ensure linearity, the power addition efficiency generally becomes low, of the order of 4%. Assuming that the output power, the power addition efficiency, the supply voltage and the power gain of the amplifier AMP2 are 8 dBmW, 4%, 3 V and 25 dB, respectively, the consumption current will become 52.4 mA. Assuming that the switches S73 and S74 are removed and the amplifier AMP3 is connected at all times, the amplifier AMP2 has only to provide an output power of 6 dBmW, requiring the consumption current to be as low as 33.1 mA. This current corresponds to 63% of that when the amplifier AMP3 is bypassed. Further, the consumption current is reduced to 78% in total when the amplifiers AMP2 and AMP3 are bypassed.
Although the switch-based bypassing method apparently seems to be simple and effective, the consumption current increases greatly with increasing output power. Thus, as a conventional manner, the bypass circuit 75 of the amplifier AMP2 is used positively while the bypass circuit 76 of the amplifier AMP3 is little used. The provision of switches also causes a problem of an increase in the number of components, the mounting area thereof, and the signal distortion.